Poverty especially deep poverty among children can have both immediate and lasting negative consequences. Children in poor families are worse off in terms of health and education, and are more likely to become teen parents. The United States has seen child poverty rates fall from record highs in the early 1990s to record lows in 2000, and many political leaders point to this decline as evidence that the 1996 federal welfare reform legislation has been a success. While poverty rates have dropped, recent "research shows that there is a group of mothers and children at the bottom of the income distribution who are worse off as a result of welfare reform" (Blank and Haskins 2001). Falling poverty rates coupled with this deterioration in economic well-being for those at the bottom of the income distribution suggest that not all groups have benefited from welfare reform and the strong economy and that more than the poverty rate is needed to understand poverty. Using longitudinal data from the Survey of Income and Program Participation (SIPP), the Panel Study of Income Dynamics (PSID), and the Urban Institute's Welfare Rules Database, this project will provide a picture of the poverty status of children and their families over the last two and a half decades and evaluate the relationship between welfare policies and poverty. To provide a picture of changes in poverty over the last 25 years, this study will examine four distinct poverty measures: (1) poverty rate, (2) deep poverty rate, (3) average income of the poor, and (4) distribution of income among the poor. The latter three take account of the degree of poverty, thereby filling in the incomplete poverty picture provided by the poverty rate alone. To evaluate the relationship between welfare policies and poverty, the multivariate framework estimates the effect of specific welfare policies on poverty and deep poverty. This relationship is estimated for both children and their mothers, by race and ethnicity. Identifying the effect of welfare reform on poverty requires disentangling the effect of state and federal welfare policies from other factors that affect poverty, such as other social policies, economic conditions, and demographic shifts in the population. This is accomplished in part by the unique design of our study population: a constant population of ever single mothers who are both comparable in the pre- and post-welfare reform periods and vulnerable to welfare receipt and poverty. In addition, the econometrics specification includes many individual- and state-level variables, as well as state, time, and individual-level fixed effects. A natural extension of this study is to carry out similar analyses using a broader measure of income--an experimental poverty measure.